Melatonin receptor ligands: Synthesis of new melatonin derivatives and comprehensive comparative molecular field analysis (CoMFA) study

Article abstract of DOI:10.1021/jm9810093

The CoMFA methodology was applied to melatonin receptor ligands in order to establish quantitative structure-affinity relationships. One hundred thirty-three compounds were considered: they were either collected from literature or newly synthesized in order to gain information about the less explored positions. To this end, various melatonin derivatives were prepared and their affinity for quail optic tecta melatonin receptor was tested. Compounds were aligned on the putative active conformation of melatonin proposed by our previously reported pharmacophore search, and their relative affinities were calculated from the displacement of 2-[¹²⁵I]-iodomelatonin on different tissues expressing aMT receptors. Compounds were grouped into three sets according to their topology. Subset A: melatonin-like compounds; subset B: N-acyl-2-amino-8-methoxytetralins and related compounds; subset C: N-acyl-phenylalkylamines and related compounds. CoMFA models were derived for each set, using the steric, electrostatic, and lipophilic fields as structural descriptors; the PLS analyses were characterized by good statistical parameters, taking into account the heterogeneity of the binding data, obtained with different experimental protocols. From the CoMFA model for the melatonin-like compounds, besides the well-known positive effect of 2-substitution, a low steric tolerance for substituents in 1, 6, and 7, and a negative effect of electron-rich 4-substituents were observed; the information provided by the newly synthesized compounds was essential for these results. Moreover, a comprehensive model for the 133 compounds, accounting for a common alignment and a common mode of interaction at the melatonin receptor, was derived (Q² = 0.769, R² = 0.905). This model validates our previously reported pharmacophore search and offers a clear depiction of the structure-affinity relationships for the melatonin receptor ligands.

Full text of DOI:10.1021/jm9810093

J . Med. Chem. 1998, 41, 3831-3844
3831
Mela ton in Recep tor Liga n d s: Syn th esis of New Mela ton in Der iva tives a n d
Com p r eh en sive Com p a r a tive Molecu la r F ield An a lysis (CoMF A) Stu d y
Marco Mor,* Silvia Rivara, Claudia Silva, Fabrizio Bordi, and Pier Vincenzo Plazzi
Dipartimento Farmaceutico, Universit a` degli Studi di Parma, viale delle Scienze, I-43100 Parma, Italy
Gilberto Spadoni, Giuseppe Diamantini, Cesarino Balsamini, and Giorgio Tarzia
Istituto di Chimica Farmaceutica e Tossicologica, Universit a` degli Studi di Urbino, piazza Rinascimento 6,
I-61029 Urbino, Italy
Franco Fraschini, Valeria Lucini, Romolo Nonno, and Bojidar Michaylov Stankov
Cattedra di Chemioterapia, Dipartimento di Farmacologia, Universit a` degli Studi di Milano, via Vanvitelli 32,
I-20129 Milano, Italy
Received March 2, 1998
The CoMFA methodology was applied to melatonin receptor ligands in order to establish
quantitative structure-affinity relationships. One hundred thirty-three compounds were
considered: they were either collected from literature or newly synthesized in order to gain
information about the less explored positions. To this end, various melatonin derivatives were
prepared and their affinity for quail optic tecta melatonin receptor was tested. Compounds
were aligned on the putative active conformation of melatonin proposed by our previously
reported pharmacophore search, and their relative affinities were calculated from the
displacement of 2-[1 I]-iodomelatonin on different tissues expressing aMT receptors. Com-
pounds were grouped into three sets according to their topology. Subset A: melatonin-like
compounds; subset B: N-acyl-2-amino-8-methoxytetralins and related compounds; subset C:
N-acyl-phenylalkylamines and related compounds. CoMFA models were derived for each set,
using the steric, electrostatic, and lipophilic fields as structural descriptors; the PLS analyses
were characterized by good statistical parameters, taking into account the heterogeneity of
the binding data, obtained with different experimental protocols. From the CoMFA model for
the melatonin-like compounds, besides the well-known positive effect of 2-substitution, a low
steric tolerance for substituents in 1, 6, and 7, and a negative effect of electron-rich
25
4
-substituents were observed; the information provided by the newly synthesized compounds
was essential for these results. Moreover, a comprehensive model for the 133 compounds,
accounting for a common alignment and a common mode of interaction at the melatonin
receptor, was derived (Q ) 0.769, R ) 0.905). This model validates our previously reported
pharmacophore search and offers a clear depiction of the structure-affinity relationships for
the melatonin receptor ligands.
2
2
In tr od u ction
named Mel1a, Mel1b, and Mel1c.11 Recent progress in
this area has been the cloning of melatonin receptors
Melatonin (N-acetyl-5-methoxytryptamine, aMT, AI-
1
2
from Xenopus dermal melanophores and from ham-
1
) is the principal hormone of the vertebrate pineal
1
3
1
ster, sheep, and human brains. The hormone has also
gland, and is produced mainly at night. Melatonin
synchronizes circadian rhythms in birds, reptiles, and
been found to have an influence on the immune system1
and to be useful as a coadjuvant in cancer therapy.
4
2
15
mammals, modulates several aspects of retinal physiol-
ogy,3 and regulates some aspects of reproduction in
Besides the receptor-mediated effects, aMT is a potent
4
16
seasonally breeding animals. aMT has been implicated
in a number of pathological states, suggesting its
therapeutic application in several disorders such as
radical scavenger, protects neurons from kainate-
1
7
induced excitotoxicity, and inhibits nitric oxide syn-
1
8
thase, suggesting that aMT or its synthetic analogues
might be considered for use as pharmacological agents
5
6
delayed sleep-phase syndrome, seasonal depression,
jet-lag, shift work disturbances,8 and as a hypnotic
7
19
for the treatment of neurodegenerative pathologies.
The field has recently been reviewed.
agent.9 These effects are achieved through the binding
20
1
0
of aMT to high affinity G-protein coupled receptors,
which have been classified into different subtypes
In the past decade, the synthesis of several potent
indole and non-indole melatonin receptor ligands has
advanced our knowledge of the structural requirements
for the binding of aMT to its receptors. However, for a
*
Corresponding author: Marco Mor. Tel: ++39 521 905063. Fax:
+
+39 521 905006. E-mail: mor@ipruniv.cce.unipr.it.
S0022-2623(98)01009-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/05/1998

3
832 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
rational drug design in this area, a model capable of
quantitatively predicting the biological activity of new
compounds would be highly desirable.
In the present paper we report the results of a new
3D-QSAR study performed with the aim of finding out
quantitative structure-affinity relationships between
different chemical classes of melatonergic ligands. At
the moment there are no data on the functional activity
of a large number of ligands; although the issue of
structure-efficacy relationships is not examined in this
work, we analyzed compounds with at least partial
agonism, or ligands having only minor structural dif-
ferences from known agonists (see Pharmacology and
Discussion). In our opinion, this allows the hypothesis
of a common alignment for the search of 3D-QSARs.
Different classes of melatonin receptor ligands have
recently been reported. They have been designed on the
basis of their bioisosterism with the indole structure of
2
1
aMT, such as the naphthalene analogues or the new
-(N-acyl-2-aminoethyl)indole derivatives,22 or with the
aim of verifying whether simplified structures, such as
1
the phenyl derivatives,2
1a,23
could maintain high binding
affinity. Moreover, many structurally different classes
of conformationally constrained compounds have been
developed,24 to test the possible relative spatial orienta-
tions of the groups interacting with the receptor.
Despite the recent classification of Mel1 receptors in
11
three subtypes (a, b, and c), a huge amount of
From qualitative and quantitative structure-affinity
literature binding data comes from tissues expressing
2
9c
studies, the 5-methoxyaryl and the amido moieties of
more than one subtype.
Although further studies
aMT have been found to be essential in achieving high
could be able to exploit different SARs for different
receptor subtypes, this is impossible with the available
binding data. The present paper refers therefore to the
affinity melatonergic ligands,2
4a,25
and various chemical
substituents on the 2 position of the indole ring to
enhance the binding affinity.26 To define possible
pharmacophore models for this receptor, we submitted
aMT and other conformationally restricted ligands to
125
potency of the ligands in displacing 2-[ I]-iodomelato-
nin from different tissues, apparently behaving in the
same way with respect to structure-affinity profiles.
This behavior has been observed in a detailed study on
the binding of 21 ligands, representative of the struc-
tural variation considered in our work, to different
2
7
the pharmacophore searching procedure DISCO; two
models (models A and B) were obtained, in which eight
putative pharmacophore points, originating from the
previously cited methoxyaryl and amido groups, were
considered. The putative active conformations of aMT
proposed by these models are characterized by the
methoxy group in the plane of the indole ring, with the
methyl group pointing toward the side chain, which in
turn is positioned orthogonally to the indole ring; the
two models differ in the orientation of the amido group
upon rotation of the CR-N bond.2
29d
tissue preparations.
The CoMFA methodology31 was applied to an ex-
tended set of compounds (133) [Figure 1, Table 1], some
known from literature and others specifically synthe-
sized, to gain information about the less explored
positions. While the effect of some substituents at
position 2 of the indole nucleus has been studied in
4b
2
6
detail, the effect of introducing substituents at other
positions has been poorly examined, if at all. Some new
derivatives were therefore synthesized (AI-12, -13, -15,
Other authors have recently proposed both 3D-QSAR
and pseudoreceptor models of the putative binding site
2
8
based on limited sets of compounds. Sicsic et al.
-35, -38, -41, -43, -44) and some known compounds (AI-
reported the results of a CoMFA study applied to a set
of compounds limited in number (48) and in chemical
diversity, and based on a different approach (see Results
and Discussion). These compounds were tested in a
binding assay on chicken brain membranes and aligned
on a conformationally constrained phenalene derivative
characterized by moderate affinity; the resulting phar-
macophore model is different from those proposed by
1
8, -19, -20, -29, -30, -31, -32, -42) were re-prepared to
be tested, to increase the information available. In
particular, the methoxy group in position 5 was substi-
tuted with halogens (AI-29, -30, -32), methyl (AI-31),
or the bulkier 2-hydroxyethyloxy group (AI-35), or
moved to other positions, 4 (AI-18), 6 (AI-19), and 7 (AI-
2
0), to test whether the effect on affinity of a variation
2
4b
in the topology of essential attachment points could be
accounted for by a topographical 3D-QSAR model. The
effect of a halogen (Br) was evaluated in position 6 (AI-
us,
and therefore the relative alignment is also
2
5
29a
30
different. Sugden, Grol, and Navajas, on the other
hand, proposed three different models of the putative
structure of the melatonin receptor. These models differ
from one another not only in the building approach, the
first two receptors being modeled on the known 3D
structure of bacteriorhodopsine and the last one on that
of rhodopsine, but also as to the putative binding points
of aMT and the nature and position of the amino acids
involved in the interaction. The papers by Sugden and
Navajas seem to derive a ligand-receptor interaction
scheme only on the basis of the putative interaction
points and of the protein primary sequence, as they
consider aMT in its extended form, and no conforma-
tional and/or pharmacophore search is reported. The
homology modeling study presented by Grol confirms
the results of the pharmacophore search reported by
3
8), and different substituents were introduced on the
indole nitrogen (AI-42, -43, -44). The simultaneous
presence of two halogen groups was evaluated in
compounds AI-12 and AI-13 while several methoxy
groups were introduced in compounds AI-15 and AI-
4
1. Compounds without an acetylamino or propiony-
lamino side chain were excluded from the analysis, as
we were not interested in studying the effects of the side
chain substituents, for which, in any case, attempts at
2
1b,23,30
explanation have already been made.
For the
same reason, we did not consider the R or â substituted
derivatives on the amidoethyl side chain. The CoMFA
methodology was applied to all the 133 melatonergic
ligands, taking as a starting point our previously
2
9b
24b
J ansen,
which proposed an active conformation of
proposed putative pharmacophore models.
Several
2
4b
aMT differing from ours
amido group.
in the orientation of the
CoMFA analyses, including molecular lipophilicity po-
3
2
tential (MLP), together with the steric and electro-

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3833
F igu r e 1. Schematic representation of compounds included in the analysis (see Table 1). The common substructures used in the
alignment are represented in bold.
static fields, were performed, grouping the melatonin
receptor ligands into three different structural classes
1000 times less potent than aMT. The introduction of
a halogen atom (AI-30, -32, and -29) led to a loss of
affinity, compared to aMT, which was limited for Br and
Cl, and higher for F; a methyl group (AI-30), or the
bulkier 2-hydroxyethyloxy group (AI-35), caused an
even greater loss of affinity. Moreover, the topology of
aMT resulted unique, as other methoxy tryptamine
derivatives displayed significantly less affinity, in
the order 4-OCH3 (AI-18) > 7-OCH3 (AI-20) > 6-OCH3
(see Data Set and Classification in the Experimental
Section). Subset A: melatonin-like compounds; subset
B: N-acyl-2-amino-8-methoxytetralins and related com-
pounds; subset C: N-acyl-phenylalkylamines and related
compounds (Table 1). A global model for the compounds
of the three classes and for unclassified ligands (subset
D) was built under the hypothesis of a common mode
of interaction at the melatonin receptor.
(
AI-19). When more than one methoxy group is pre-
sent, the binding of the 5- group seems prevalent, the
,7-dimethoxy derivative (AI-41) having a higher affin-
Resu lts a n d Discu ssion
5
ity than the 7-methoxy; compound AI-41 provides
unique information on the 7 position of the indole ring.
Alkylation at the nuclear nitrogen leads to a slight
decrease in affinity with a small methyl group (AI-42),
and to a more pronounced one with bulkier groups (AI-
The results of binding studies for newly tested com-
pounds, reported in Table 2, gave new information on
SAR for melatonergic ligands. The positive effect of a
2
-Br substitution, already observed for compound AI-5
(Table 1), was partially reversed by the introduction of
4
3, -44).
an additional Br atom in position 6 (AI-13), and fully
reversed when the second halogen was introduced in
position 4 (AI-12). Negative effects for substitution in
positions 4 and 6, ortho to the 5-methoxy group, were
also observed for compounds AI-15 and AI-38.
The absence of the 5-methoxy group gave a general
drop in affinity, as expected from previous SAR, the
known N-acetyltryptamine (AI-16 in Table 1) being
The 133 melatonin receptor ligands reported in Table
1 were aligned on the aMT conformation corresponding
to our pharmacophore model B (see Alignment Rules
in the Experimental Section) and grouped in three
subsets (A, B, and C; see Data Set and Classification)
which were submitted to CoMFA analysis; the results
of the 3D-QSAR analyses for each subset and for the

3
834 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
Ta ble 1. Melatonin Receptor Ligands Included in CoMFA Analyses (see Figure 1 for general formulas and label definition)
pRA pRA calcd
pRAb ref calcd in subset
c
d
compd
R
R1
R2
R4
R5
R6
R7
X
AI-1
AI-2
AI-3
AI-4
AI-5
AI-6
AI-7
AI-8
AI-9
AI-10 CH3
AI-11 CH3
AI-12 CH3
AI-13 CH3
AI-14 CH3
AI-15 CH3
AI-16 CH3
AI-17 CH2CH3
AI-18 CH3
AI-19 CH3
AI-20 CH3
AI-21 CH3
AI-22 CH3
AI-23 CH2CH3
AI-24 CH3
AI-25 CH3
AI-26 CH3
AI-27 CH3
AI-28 CH2CH3
AI-29 CH3
AI-30 CH3
AI-31 CH3
AI-32 CH3
AI-33 CH3
AI-34 CH3
AI-35 CH3
AI-36 CH3
AI-37 CH3
AI-38 CH3
AI-39 CH3
AI-40 CH3
AI-41 CH3
AI-42 CH3
AI-43 CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH3
CH3
CH3
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
C6H5
C6H5
Br
H
H
H
H
H
H
H
H
H
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
Br
Cl
OCH3
H
H
H
OCH3
H
F
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
N
0.00
-0.71
-0.47
-0.18
0.92
1.25
0.95
1.12
0.71
0.63
0.13
-0.18
-1.51
-0.80
0.73
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
O
S
0.33 24c -0.30
1.30 26
1.10 51
1.30 26
1.70 26
1.00 25
0.41 26
-0.59 26
0.29
0.72
0.74
0.95
0.48
0.23
0.05
I
Cl
CH3
CH(CH2)2
C6H11
CH2C6H5
Br
Br
CH3
H
H
H
H
H
H
H
C6H5
C6H5
C6H5
C6H5
Br
Br
Br
H
H
H
H
H
H
H
H
H
H
H
-0.68 26 -0.38
-1.50 60 -1.67
-1.07
0.42
0.40 24c
-3.09
-3.07 52 -3.06
-2.83 52 -2.68
-2.52
-3.40
-3.13
-2.37 28 -2.50
-2.23 51 -2.13
-2.07 51 -1.70
-2.03 51 -2.45
-2.42 51 -3.08
-2.37 52 -1.70
-2.32 52 -1.63
-1.66 52 -1.25
a
Br
H
H
-0.83
0.81
0.51
-3.43
a
0.34
a
OCH3 OCH3
-3.74
-3.22
-2.95
e
e
e
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
Br
CH3
Cl
a
-2.35
-3.22
-3.13
a
a
-2.75
-1.87
-1.57
-2.16
-2.88
-1.82
-2.03
-1.85
-2.61
-2.09
-2.26
-2.22
-2.20
-3.09
-2.52
0.02
-0.60
-0.71
-0.79
-1.94
-2.49
-0.65
-2.34
-2.68
-0.89
-0.50
-1.87
-0.62
-0.32
1.03
CH3
CH2CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Br
Br
H
H
H
H
H
H
H
a
-2.13
-1.45
-2.61
-1.70
-2.65
-2.25
-2.17
-2.32
a
a
a
OH
-3.31 25 -2.12
-2.85 25 -2.87
OCH2C6H5
O(CH2) 2OH
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
H
OCH3
OCH3
OCH3
OCH3
OCH3
H
a
-2.69
-2.45
F
-0.18 25 -0.11
-0.30 26 -0.48
Cl
Br
OH
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
OCH3
OCH3
H
H
H
H
OCH3
H
a
-0.70
-0.62
-1.42 25 -0.93
-2.12 25 -1.99
H
H
H
H
H
H
H
H
H
C6H5
H
OCH3
H
H
H
a
-2.32
-1.04
-2.66
-2.92
-2.33
-0.90
-2.48
-2.49
a
CH3
CH2C6H5
C6H5
a
a
AI-44 CH3
AI-45 CH3
AI-46 CH3
H
H
-0.97 21b -1.15
-0.79 21b -0.50
-1.58 22 -1.80
-0.74 22 -0.61
-0.51 22 -0.17
1.15 22
0.42 22
1.70 22
-3.39 22 -2.82
-2.22 22 -2.38
0.09 21a
0.57 28
0.82 21a
1.00 21a
AII-1
AII-2
AII-3
AII-4
AII-5
AII-6
AII-7
AII-8
AIII-1 CH3
AIII-2 CH2CH3
AIII-3 CH3
AIII-4 CH2CH3
AIII-5 CH3
AIII-6 CH3
CH2CH3
CH3
H
H
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
Br
COOCH3
C6H5
H
I
H
H
OCH3
OCH3
H
H
H
1.24
1.03
0.74
0.35
0.97
e
e
-0.04
0.26
e
e
H
OCH3
OCH3
OCH3
OCH3
H
0.03
0.47
0.98
1.38
-2.73 21b -2.74
-2.68 21a -2.44
-1.65 28 -1.47
-0.93 24e -0.82
-2.18 24e -1.43
-0.18 24e -0.23
e
H
OH
-2.80
-1.75
-1.02
-1.39
-0.24
-0.19
-2.23
-1.27
-2.31
-2.23
-3.15
-2.23
f
AIII-7 CH3
AIV-1 A ) furan
H
AIV-2 A ) 4-oxo-4,5-dihydrofuran
AIV-3 A ) pyran
AIV-4 A ) 2H-3,4-dihydropyran
-0.37 24e
0.08
BI-1
BI-2
BI-3
BI-4
BI-5
BI-6
BI-7
BI-8
BI-9
CH3
CH2CH3
CH3
CH3
CH3
CH2CH3
CH3
CH3
OCH3
OCH3
OCH2CH3
OCH3
H
H
H
H
OCH3
OCH3
H
H
H
H
H
H
OCH3
H
H
H
H
H
Cl
H
H
H
OCH3
H
H
CH2 -1.91 24c -2.61
CH2 -1.11 24c -1.89
CH2 -2.53f 29b -2.39
CH2 -2.23f 29b -2.29
CH2 -3.07 24c -3.41
CH2 -2.41 24c -2.77
CH2 -2.33 24c -2.53
CH2 -3.53 24c -3.17
f
CH3
O
O
-4.16 24d -3.14
-2.89 24d -2.44
-3.95
-2.94
BI-10 CH2CH3
H

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3835
Ta ble 1. continued
pRA
pRA calcd
compd
R
R1
R2
R4
R5
R6
R7
X
pRAb
ref calcdc in subset
d
BII-1
BII-2
BII-3
BII-4
BII-5
BIII-1
BIII-2
CH3
CH2CH3
CH3
CH3
CH3
OCH3
OCH3
OCH3
OCH3
H
H
H
OH
OCH3
H
-1.71
-1.12
-1.84
-0.18
-2.63
-2.74
-1.52
-2.77
-2.56
-2.99
-2.36
-2.02
-1.42
-1.80
-1.08
-2.46
-1.98
-2.51
-2.63
-3.21
-2.02
-3.17
-2.34
-3.38
-2.80
-2.03
-0.98
-3.51
-3.10
-3.03
-3.16
-2.73
-2.76
-2.30
-0.60
0.00
28 -1.66
-1.78
-0.84
-1.82
-0.17
-2.84
-2.54
-1.62
-2.82
-2.66
-2.73
-1.94
-2.19
-1.40
-1.97
-1.18
-2.63
-1.83
-2.39
-2.68
-2.68
-2.06
-3.12
-2.51
-3.41
-3.07
-1.85
-1.21
-3.57
-2.93
-3.32
-3.28
-2.64
-2.55
-2.38
-0.98
-0.35
-0.66
-0.02
-0.91
-2.86
24f -0.95
24f -1.26
24f -1.09
24f -2.86
24b -2.10
24b -1.21
24f -3.20
24f -2.83
H
Br
BIV-1 A ) benzo
BIV-2 B ) benzo
CI-1
CI-2
CI-3
CI-4
CI-5
CI-6
CI-7
CI-8
CH3
CH2CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
CH2CH3
CH3
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OCH3
OCH3
Br
Br
CH3
CH3
CH2CH3
C6H5
H
H
H
H
H
H
H
H
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
4
4
23
23
-2.90
-2.20
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
H
H
H
H
OCH3
OCH3
H
H
H
H
H
H
H
OCH3
H
OCH3
OCH3
OCH2CH3
OCH2CH3
OCH3
21a -2.03
21a -1.33
21a -2.45
21a -1.74
21a -2.51
28
-1.80
CI-9
21a -2.16
CI-10
CI-11
CI-12
CI-13
CI-14
CI-15
CI-16
CI-17
CI-18
CI-19
CI-20
CI-21
CI-22
CI-23
CI-24
CI-25
CII-1
CII-2
CII-3
CII-4
CII-5
CIII-1
DI-1
28
23
23
23
23
23
23
23
23
23
23
23
23
23
23
23
-3.42
-2.56
-2.05
-3.42
-2.91
-3.32
-2.96
-1.57
-1.07
-3.62
-3.11
-3.29
-3.20
-2.69
-2.42
-1.99
OCH3
OCH3
OCH3 OCH3
OCH3 OCH3
H
H
OCH3
OCH3
F
F
Cl
Br
Br
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
CH2CH3
CH2CH3
CH2CH3
CH3
CH2CH3
CH3
H
OCH3
H
H
H
H
21a -0.76
21a -0.23
-0.72
-0.20
-1.25
-2.77
-1.12
0.19
28
28
28
-0.61
-0.08
-1.06
CH2CH3
CH2CH3
OCH3
24b -2.06
24b -1.07
H
C6H5
DI-2
24b
0.12
DII-1
DII-2
DIII-1 CH3
DIII-2 CH3
DIII-3 CH3
DIII-4 CH2CH3
DIII-5 CH3
DIII-6 CH2CH3
DIV-1 CH3
DIV-2 CH3
DIV-3 CH2CH3
DIV-4 CH2CH3
H
-0.65
24b -1.35
24b -0.22
24b -1.62
24b -3.16
24a -2.15
24a -1.85
24a -3.57
24a -3.23
24a -0.43
24a -2.85
24a -0.06
24a -2.46
24g -2.10
24g -1.40
COOCH2CH3
0.11
OCH3 A ) benzo
H A ) benzo
-1.80
-2.91
-2.57
-1.84
-3.96
-2.87
-0.21
-2.59
-0.39
-2.54
-2.12
-1.22
OCH3 A ) tetrahydrobenzo
OCH3 A ) tetrahydrobenzo
H
H
OCH3
H
A ) tetrahydrobenzo
A ) tetrahydrobenzo
OCH3
H
DV-1
DV-2
DV-3
DVI-1
CH3
CH2CH3
CH2CH3
1
1
0
-2.83g 24g -2.79
-1.76 24g -1.84
a
Compounds newly synthesized or tested. b Negative logarithm of relative affinity of compounds, compared to that of aMT in the same
c
d
experiment, and used as the dependent variable in the CoMFA analyses. pRA calculated by the comprehensive model. pRA calculated
e
f
by the best CoMFA analysis for each subset. Excluded by the partial model as topologically different from aMT. Excluded by the partial
model as topologically different from the other N-acyl-2-amino-8-methoxytetralins. g Negative logarithm of the harmonic mean calculated
from the values of the enantiomers (see details in the Experimental Section).
comprehensive model are summarized in Table 3. The
best analyses were selected on the basis of their predic-
tive power.
Subset A includes those melatonin receptor ligands
identical from a topological point of view to the natural
ligand. The best model is a four latent variable (LV)
steric and electrostatic one, whose graphical representa-
tion is shown in Figure 2 (top). The most important
positive steric regions (green) are situated near aMT
position 2, where a substituent is considered to be able
to enhance the binding affinity, and around the CH3 of
the 5-methoxy group. The latter allows us to distin-
guish between the majority of methoxy derivatives and
those having the methyl group perpendicular to the
aromatic ring (because of the presence of 4-substitu-
ents), which are generally less potent. The negative
steric region (red) corresponding to position 6, 7 of aMT,
shows that affinity is not favored by the presence of

3
836 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
Ta ble 2. Binding Affinitya of Newly Tested Melatonin Analogues for the Melatonin Receptor
compd
R1
R2
R4
R5
OCH3
OCH3
OCH3
H
H
H
F
Br
R6
R7
IC50
Ki
RAb
aMT (AI-1)
AI-12
AI-13
AI-15
AI-18
AI-19
AI-20
AI-29
AI-30
AI-31
AI-32
AI-35
AI-38
AI-41
AI-42
AI-43
AI-44
2.2
23.3
0.61
5.75
0.176
1
11.2
0.38
1234
332
2517
1360
136
28
405
50
492
5
H
H
H
H
H
H
H
H
H
H
H
H
H
Br
Br
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Br
H
H
H
H
H
H
H
Br
OCH3
H
OCH3
H
0.654
2460
772
4800
2620
295
33.6
719
126
OCH3
OCH3
H
617
190
1180
647
72.8
8.91
180
30
289
2.98
121
6.72
287
380
H
H
H
H
H
H
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
CH3
Cl
O(CH2)2OH
OCH3
OCH3
OCH3
OCH3
OCH3
H
1020
Br
H
12.8
491
25.5
1090
1470
OCH3
H
211
11
456
828
CH3
CH2C6H5
C6H5
H
H
H
H
H
H
H
H
a
IC50 and Ki values are expressed in nM and are the means of at least three independent determinations performed in duplicate,
derived from nonlinear fitting strategies. The SEM values were below 15% of the mean. b Relative affinity ) [(IC50 compd.)/(IC50 aMT)]
determined in parallel, in the same experiment.
Ta ble 3. Statistics of the CoMFA Models
data set
N
fields
S, E
LVs
4
R2
s
Q2 a
SDEP a,b
0.701
subset A: melatonin-like compounds
57
0.921
0.408
0.745
%S 74.4
E 25.6
%S 46.6
L 36.9
E 16.5
%S 67.9
E 32.1
%S 68.4
E 31.6
%
subset B: N-acyl-2-amino-8-methoxytetralins
17
S, L, E
3
0.965
0.189
0.692
0.495
%
%
subset C: N-acylphenylalkylamines
31
S, E
S, E
3
5
0.948
0.905
0.232
0.421
0.785
0.769
0.440
0.639
%
global set
133
%
a
2
With leave one out procedure; with four cross-validation groups and the reported number of latent variables: Q ) 0.742 (subset A),
b
2
1/2
0
.619 (subset B), 0.810 (subset C), 0.736 (global set). Calculated as [Σ(y-yPRED) /N] (ref 61).
substituents in this area. Another negative steric region
is observed between positions 1 and 2 of aMT, above
the indole plane; this is essentially due to the presence
in this region of the two benzyl groups of compounds
AI-11 and AI-43, both less potent than aMT. The
electrostatic field contributes to a lesser extent (25.6%
vs 74.4%) to the explanation of the variation in affinity.
Two magenta regions (negative electrostatic potential
favorable) are observed close to the aMT methoxy group.
The one to the left of the oxygen atom indicates that
this atom probably binds to the active site through some
electrostatic interaction or hydrogen bond. The second
magenta region, inscribed into the green (steric) one, is
due to the 5-halogen-substituted derivatives and to
those compounds having the 5-methoxy group out of the
indole plane, which are more potent than the 5-hydrogen-
substituted compounds. In fact, when 5-halo-deriva-
tives are excluded from the analysis, this region becomes
less pronounced, but it is still present. Another elec-
trostatic region (positive potential favorable, black) is
positioned near the indole NH group of aMT and is due
to the benzofuran and benzothiophene derivatives AI-
erogeneity of the binding data, obtained with different
experimental protocols, prevents further improvement
of the statistics of the model; this is not the case for
subsets B and C, where more homogeneous values were
available.
The N-acyl-2-amino-8-methoxytetralins and related
compounds in subset B are all less potent than aMT.
This is probably due to the constrained spatial disposi-
tion of the melatonin-like amidoethyl side chain and/or
to the methoxy group orientation out of the plane of the
ring, which could make interaction with the receptor
more difficult. The presence of a third phenyl ring,
condensed with the tetralin structure, seems to improve
potency, as can be seen in the tetrahydrophenalene
derivatives BII. The best model obtained for subset B
was a three latent variable PLS analysis in which the
lipophilic, steric, and electrostatic fields were correlated
with the affinity for the melatonin receptor. The
graphical representation (Figure 2, center) shows a
magenta region near the methoxy oxygen, where the
presence of negative electrostatic potential can enhance
affinity. Positive steric (green) and lipophilic (yellow)
regions are present near positions 4 and 5 of the
tetralins, surrounding the third ring of the BII com-
pounds cited. The black region inscribed therein is due
4
5 and AI-46. The PLS analysis obtained for subset A,
despite a residual standard error of 0.41 log units,
accounts for 92.1% of the affinity variation. The het-

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3837
F igu r e 2. CoMFA stdev*coeff. contour plots for subset A (top, contour level: 0.006 for steric and 0.004 for electrostatic fields),
subset B (center, contour levels: 0.006 for steric and lipophilic, 0.002 for electrostatic fields), and subset C (bottom, contour
levels: 0.006 for steric and 0.002 for electrostatic fields), according to the PLS models described in Table 3. Color codes - green:
steric positive; red: steric negative; black: electrostatic positive; magenta: electrostatic negative; yellow: lipophilic positive;
cyan: lipophilic negative.
to the chromane derivatives BI-9, -10, whose electrone-
gative oxygen atom exerts a deep negative effect on
affinity.
Subset C is composed of 31 compounds characterized
by an N-acylaminoalkyl side chain and a methoxy group
As can be seen in Figure 2 (bottom), the most important
regions are positive steric (green), which correspond to
the methoxy group, the condensed ring in naphthalene
and indole derivatives, and the acylamino chain. The
electrostatic contribution, visible only at lower coef-
ficient levels, is mainly due to the methoxy group, whose
presence seems to exert considerable influence on af-
finity for the melatonin receptor also in this subset.
However, the effect of the negative charge on the oxygen
atom is not observed because of the presence, in this
region, of electron-rich groups only. The positive partial
charges on the methyl group cause scattered black
regions accounting for the lower affinities of the halogen-
substituted compounds (CI-19-CI-23). At the low
coefficient level applied to the electrostatic field, a
(when present) bound to the same phenyl ring. It was
not possible to obtain a very close alignment for subset
C, as these compounds have ethylamido, propylamido,
or butylamido chains, and the methoxy group positioned
in ortho or meta to the side chain. In addition some
naphthalene and indole derivatives were included (CII
and CIII), topologically similar to other compounds in
this group. We obtained a three latent variable PLS
model in which the steric contribution is greater than
the electrostatic one (67.9% and 32.1%, respectively).

3
838 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
F igu r e 4. Representation of the MOPAC minimum-energy
conformation of a hydroxydihydrofuran derivative of aMT
endowed with low affinity (see ref 24e), indicating the in-
tramolecular hydrogen bond.
were aligned in a topographical way, fitting the hypo-
thetical attachment points (see Experimental Section).
Some problems of ambiguity in the alignment were also
resolved by this model: two naphthalene compounds
(AIII-3 and AIII-4) having two methoxy groups (in 2
and 7 positions) could be aligned as in subset A, with
the 7-OCH3 on the 5-OCH3 of aMT, or as belonging to
subset C, having the 2-OCH3 on the same ring as the
side chain; they were therefore excluded during the
building of partial models. The prediction of their
affinity by the comprehensive model gave good results
when they were aligned in the first way (pRA obsd:
AIII-3 ) 0.82, AIII-4 ) 1.00; pRA pred ) 0.62 and 1.03,
respectively), but not in the other case (pRA pred )
-0.86 and -0.29).
F igu r e 3. Comprehensive CoMFA model for the global set of
33 compounds: steric and electrostatic stdev*coeff. contour
plots (contour level 0.006) surrounding aMT in capped sticks.
Color codes - green: steric positive; red: steric negative;
gray: electrostatic positive; magenta: electrostatic negative.
1
magenta region (negative electrostatic potential favor-
able) also appears to the right of the phenyl ring; this
is mainly due to the positive effect on affinity caused
by the Br substituents in compounds CI-5 and CI-6.
The three models described so far give an accurate
explanation of the 3D quantitative structure-affinity
relationships for each subset of melatonin receptor
ligands. We also derived a comprehensive model in-
cluding compounds from subsets A, B, and C, and all
those compounds which were excluded (subset D, Table
Moreover, this model supports a hypothesis of inter-
action with the receptor at the same attachment points
for all these known ligands; while within each subset
the structure-affinity profiles are quite similar (see
Figure 2), the differences among classes are accounted
for by differences in the superposition space and align-
ment, as illustrated by the scattered red (negative steric)
regions around the side chain in Figure 3 and by the
methoxy orientation discussed above. This is a proof
of the consistency of our previous pharmacophore model
1
2
), some of which were quite potent (e.g. compounds DI-
, DII-2, DIV-1, DIV-3). The best model built by PLS
2
4b
analysis with the global set was a five latent variable
steric and electrostatic model (Table 3, fourth row). The
affinity values calculated using this model are reported
in Table 1, as well as the values obtained with the best
analysis for each subset. Figure 3 depicts the most
important regions of space associated with the variation
in potency at the melatonin receptor for the global set
of ligands. As for the steric potential, it is possible to
notice a positive effect in the region corresponding to
(model B ), as all the ligands endowed with a certain
potency at the melatonin receptor could be aligned on
the putative aMT active conformation.
As for the choice of the amido group orientation (Câ-
CR-N-C3 ) τ3 was ≈180° in our pharmacophore model
2
4b
B, as opposed to ≈90° in model A ), the unexpected
low affinity reported for the condensed hydroxydihy-
drofuran derivative represented in Figure 4 (Ki 1.30 ×
-
7
24e
10
M
), compared to the good affinity of other furan
2
-substitution of aMT (green), while substituents in
and pyran derivatives (AIV), suggested to us that model
B was preferable for the extensive CoMFA study. In
fact, in this compound a hydrogen bond is possible
between the amido CO and the OH on the furan ring.
The resulting conformation, corresponding to the ori-
entation of CONH as in model A, had a minimum
energy at MOPAC calculation (ver. 6.0 implemented in
position 6 and 7 cause a decrease in affinity (red region);
apart from this raw information, no further differentia-
tion on 2-substituents was possible: a well-spread set
of derivatives tested in the same experimental condi-
tions is needed for this. The 5-methoxy group is
characterized by a steric positive interaction and by an
electrostatic interaction due to the oxygen atom (green
and magenta regions, respectively), as discussed before.
The green positive steric region near the N-acyl group
of the side chain is present in the comprehensive model,
as well as in those obtained for subsets A, B and C; since
we considered only propionyl- or acetyl-substituted
compounds, the former derivatives generally being more
active than the latter, this is only a confirmation of the
well-known acyl chain length effect.2
3
3
SYBYL, PM3 Hamiltonian with geometry optimization
and MMOK, PRECISE keywords), with an OH‚‚‚OdC
distance of 1.81 Å. As this hydrogen bond should
stabilize the conformation represented in model A, we
attributed the low affinity observed for this compound
to a bad fit of this conformation at the receptor site,
rather than to other unexplained effects of the OH
group. A CoMFA model of the global set of 133 ligands
aligned on the aMT conformation of model A (τ3 ≈ 90°)
gave the same qualitative results (CoMFA regions) with
slightly worse statistics (for a five latent variable steric
1b,30
The comprehensive model was able to explain the
rank order of affinity of the topologically different aMT
analogues (AI-18: 4-methoxy-N-acetyltryptamine, AI-
2
2
and electrostatic model: Q ) 0.731, SDEP ) 0.690, R
1
9: 6-methoxy-, AI-20: 7-methoxy-), provided that they
) 0.872, s ) 0.487).

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3839
Sch em e 1a
There is a third minimum energy orientation of the
torsion angle τ3 (≈-90°), which corresponds to one of
the putative aMT active conformations proposed by
2
9b
J ansen.
We tested the reliability of this conformation
by aligning on it the compounds used for our model; for
subsets A, B, and C, and for the global set, we obtained
CoMFA models that, although worse than ours from a
statistical point of view, could not allow us to reject this
conformation as the putative active one (for a 5 LV steric
2
2
and electrostatic model, Q ) 0.741, SDEP ) 0.677, R
0.866, s ) 0.499).
Besides the orientation of the amido group, the
)
chirality of the model remains uncertain, owing to the
lack of information about the enantioselectivity of chiral
compounds.
Our model can be compared to the CoMFA models
2
8
presented by Sicsic et al., which are based on a
different approach. These were built, however, from a
limited set of 48 compounds, most of which differ only
in the nature of the N-acyl substituent; they were tested
on the same assay and superposed onto a tetrahydro-
phenalene derivative (BII-1), both in its axial and
equatorial conformation, with no subset classification.
Our models were obtained from an extensive set ideally
containing all the information available in the literature
and from some additional information provided by the
newly tested compounds. Moreover, as it is our opinion
that statistical results of CoMFA models (both in fitting
and in leave-one-out cross-validation) are too prone to
chance correlation, to allow the choice among alternative
models in the absence of clear-cut differences, our
models were based on a pharmacophore hypothesis
a
Reagents: (a) 1-(dimethylamino)-2-nitroethylene, TFA, 0 °C,
0
.5 h; (b) POCl3, DMF, (CH2Cl)2, reflux, 0.5 h; (c) CH3NO2,
AcONH4, reflux, 1.5 h; (d) LiAlH4, THF, room temperature, 5 h;
(e) Ac O, THF, TEA, room temperature, 6 h.
2
known melatonin receptor ligands is proposed. This
model provides useful information about the structure-
affinity relationships of the whole set of compounds and
offers an indirect validation of our pharmacophore
model.
Exp er im en ta l Section
Ch em ica l Meth od s. Melting points were determined on
a B u¨ chi SMP-510 capillary melting point apparatus and are
uncorrected. H NMR spectra were recorded on a Bruker AC
00 spectrometer; chemical shifts (δ scale) are reported in parts
per million (ppm) relative to the central peak of the solvent.
EI-MS spectra (70 eV) were taken on a Fisons Trio 1000. Only
molecular ions (M ) and base peaks are given. Infrared
1
2
previously derived from several constrained analogues
of aMT.2
4b
The results of the two approaches are
+
different as to both the position of the putative attach-
ment groups and the nature and shape of the regions
of interest.
spectra were obtained on a Bruker FT-48 spectrometer;
-
1
absorbances are reported in ν (cm ). Elemental analyses for
C, H, and N were performed on a Carlo Erba analyzer.
Ch em istr y. Compounds AI-15, -19, -20, -31, -32, -35, and
The correlation between structure and functional
activity is beyond the scope of our work; some of the 3D
properties discussed here could be important for recep-
tor activation, but sufficient data for exploiting struc-
ture-efficacy relationships are not available at the
moment. The receptor subtypes may show different
SAR profiles, but these cannot be recognized from
binding data on native tissues.29d However, the align-
ment that is proposed accounts for differences in affin-
ity, and it could be useful for the exploration of 3D-
QSAR on different receptor-subtypes, or of structure-
activation relationships, as sufficient data become
available.
-
2
38 were obtained by acetylation (Ac O/TEA/THF/room tem-
perature) of the corresponding tryptamines which were com-
mercially available (6-methoxytryptamine) or synthesized by
classical methods. aMT analogues AI-18, -29, -30, and -41
were prepared by the synthetic route shown in Scheme 1 in
accordance with the method previously reported by us for
the synthesis of compounds AI-3, -8, -9, and -10. Briefly,
tryptamines were prepared by LiAlH
(
suitable indoles with 1-(dimethylamino)-2-nitroethylene or
through Knoevenagel condensation of the corresponding formyl
derivatives with nitromethane.
2
6
4
reduction of the (E)-3-
2-nitroethenyl)indole derivatives obtained by coupling of the
2,6-Dibromomelatonin (AI-13) and 2,4-dibromomelatonin
AI-12) were prepared by direct bromination of 6-bromome-
(
latonin (AI-38) and melatonin (AI-1) respectively, with 1 or 2
equiv of N-bromosuccinimide. 2-Bromomelatonin (AI-5) was
prepared as previously described.34 Compounds AI-42 and AI-
Con clu sion s
4
3 were prepared by N-alkylation of AI-1 with sodium hydride
and MeI or benzyl chloride, respectively, in DMF (Scheme 2).
The Ullmann reaction (iodobenzene, CuI, K CO , ZnO, NMP
55 °C, 6 h) was utilized for the synthesis of AI-44 (Scheme
).
Syn t h esis of 3-(2-Nit r oet h en yl)-1H -in d ole Der iva -
The CoMFA methodology was applied to a wide set
of structurally different melatonin receptor ligands in
order to derive 3D-QSAR models correlating the differ-
ences in affinity with the variation of the 3D fields. The
compounds were aligned on the putative active confor-
mation of aMT obtained from our previous pharma-
cophore search. For each class with topological homo-
geneity (subsets A, B, and C) in which the melatonin
receptor ligands were grouped, a 3D-QSAR model with
good predictive and descriptive power was obtained. In
addition, a comprehensive model suggesting a common
alignment and binding interaction mode for all the
2
3
1
2
tives: Typ ica l P r oced u r es. (E)-5-Br om o-3-(2-n itr oeth e-
n yl)-1H-in d ole. 5-Bromoindole (0.98 g, 5 mmol) was added
to a stirred ice-cooled solution of 1-(dimethylamino)-2-nitro-
ethylene (0.58 g, 5 mmol) in trifluoroacetic acid (5 mL).
The mixture was stirred at room temperature under N
.5 h and then poured onto ice-water. The aqueous solution
was extracted with ethyl acetate, and the combined organic
layers were washed with a saturated NaHCO solution and
2
for
0
3

3
840 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
Sch em e 2a
were found to be in accordance with the assigned structures
and, when available, with those reported in the literature.
Acyla tion of Tr yp ta m in e Der iva tives: Gen er a l P r o-
ced u r e. TEA (1.1 equiv) and Ac O (1.1 equiv) were added to
2
a cold solution of the suitable primary tryptamine (1 mmol)
in THF (4 mL), and the resulting reaction mixture was left
stirring at room temperature for 6 h. The solvent was
evaporated under reduced pressure, and the residue was taken
up in ethyl acetate and washed with a saturated aqueous
solution of NaHCO
Na SO , the solvent was evaporated under reduced pressure
to give a crude product, which was purified by chromatography
silica gel; ethyl acetate/cyclohexane, 7:3, as eluent) and/or
3
and then with brine. After drying over
2
4
(
crystallization.
N-Acetyl-5-flu or otr yp ta m in e (AI-29) was obtained ac-
cording to the procedure described above: 39% yield from 3-(2-
3
6
nitroethenyl)-5-fluoroindole; mp 125-126 °C (EtOAc), lit.
+
1
colorless gum; EIMS: m/z 220 (M ), 161 (100); H NMR
CDCl ): δ 1.95 (s, 3H), 2.93 (t, 2H), 3.59 (q, 2H), 5.54 (br s,
1H), 6.91-7.01 (ddd, 1H, J ) 2.53 Hz), 7.21 (d, 1H, J ) 2.36
Hz), 7.29 (m, 2H), 8.14 (br s, 1H); IR (cm , Nujol): 3269, 1635.
(
3
a
Reagents: (a) NBS, AcOH, room temperature, 4 h; (b) NaH,
-1
DMF, MeI or PhCH2Cl, room temperature, 16 h; (c) PhI, CuI, ZnO,
K2CO3, NMP 155 °C, 6 h.
N-Acetyl-5-br om otr yp ta m in e (AI-30) was obtained ac-
cording to the procedure described above starting from 3-(2-
water. After drying over Na
2
SO
4
, the solvent was evaporated
nitroethenyl)-5-bromoindole: 59% yield; mp 153-154 °C (dichlo-
romethane/hexane), lit.37 mp 153 °C. H NMR (acetone-d
1
):
under reduced pressure to give a crude product which was
purified by chromatography (silica gel; dichloromethane as
eluent) followed by crystallization from dichloromethane-
hexane. Yield (0.67 g, 50%). Chemical physical data were
6
δ 1.85 (s, 3H), 2.88 (t, 2H), 3.45 (q, 2H), 7.15 (br s, 1H), 7.20
(dd, 1H, J ) 1.8 Hz and J ) 8.6 Hz), 7.22 (d, 1H, J ) 2.04
Hz), 7.35 (d, 1H, J ) 8.6 Hz), 7.76 (d, 1H, J ) 1.78 Hz), 10.24
3
5
+
-1
identical to those reported in the literature.
-F lu or o-1H-in d ole-3-ca r boxa ld eh yd e. POCl
9 mmol) was added dropwise under nitrogen to a solution of
(br s, 1H); EIMS: m/z 280, 282 (M ), 221, 223 (100); IR (cm
,
5
3
(1.8 mL,
Nujol): 3296, 3211, 1616.
1
N-Acet yl-4,5,6-t r im et h oxyt r yp t a m in e (AI-15). From
3
8
dry DMF (2.4 mL), and the mixture was stirred at room
temperature for 15 min. Dry 1,2-dichloroethane (15 mL) was
then added and the solution cooled to -10 °C. 5-Fluoroindole
4,5,6-trimethoxytryptamine: 78% yield, oil; EIMS: m/z 292
+
1
(M ), 233 (100); H NMR (CDCl
3
): δ 1.89 (s, 3H), 2.99 (t, 2H),
3.53 (q, 2H), 3.87 (s, 6H), 4.04 (s, 3H), 6.37 (br s, 1H), 6.64 (s,
1H), 6.81 (d, 1H, J ) 2.22 Hz), 8.34 (br s, 1H); IR (cm , neat):
3323, 2934, 1628. Anal. (C15 ‚H O) C, H, N.
-
1
(1.16 g, 8.5 mmol) was added in small portions at such a rate
that the temperature did not rise above 5 °C. Finally, 3.3 g of
finely divided calcium carbonate was added and the cooling
terminated. The mixture was rapidly heated to reflux, with
mechanical stirring, and maintained at this temperature for
H
20
N
2
O
4
2
N-Acetyl-4-m eth oxytr yp ta m in e (AI-18). From 4-meth-
39
oxytryptamine: yield, and chemical-physical data were
39
identical with those reported in the literature.
3
0 min. The reaction mixture was cooled and poured into a
N-Acetyl-7-m eth oxytr yp ta m in e (AI-20). From 7-meth-
cooled solution of sodium acetate (12.5 g) in water (20 mL)
and stirred at room temperature for 2 h. After filtering, the
layers were separated and the aqueous phase was extracted
39
oxytryptamine: yield and chemical-physical data were iden-
39
tical with those reported in the literature.
N-[2-(5,7-Dim eth oxy-1H-in dol-3-yl)eth yl]acetam ide (AI-
(
2×) with CH
2
Cl
2
; the organic layers were washed with brine,
4
1). From 5,7-dimethoxytryptamine:40 79% yield, mp 167 °C
dried (Na SO
2
4
), and concentrated at reduced pressure to give
+
1
(
dichloromethane/ether); EIMS: m/z 262 (M ), 190 (100); H
NMR (CDCl ): δ 1.93 (s, 3H), 2.93 (t, 2H), 3.59 (q, 2H), 3.86
s, 3H), 3.93 (s, 3H), 5.53 (br s, 1H), 6.37 (d, 1H, J ) 1.71 Hz),
a crude residue which was filtered on silica gel (ethyl acetate/
3
cyclohexane, 1:1, as eluent) to give a crude product (1.08 g,
(
6
+
7
8%); EIMS: m/z 163 (M , 100) which was not further purified,
.62 (d, 1H, J ) 1.71 Hz), 6.99 (d, 1H, J ) 2.14 Hz), 8.13 (br
but condensed directly with nitromethane as described below.
E)-5-F lu or o-3-(2-n itr oeth en yl)-1H-in d ole. A solution of
-fluoroindole-3-carboxaldehyde (0.56 g, 3.42 mmol) and am-
-
1
s, 1H); IR (cm , Nujol): 3393, 3280, 3102, 1621.
(
N-Acetyl-5-Meth yltr yp ta m in e (AI-31). From 5-methyl-
5
tryptamine:41 73% yield, yellow oil; H NMR (CDCl
(
1
): δ 1.96
3
monium acetate (0.12 g) in nitromethane (4.8 mL) was heated
at reflux for 1.5 h under nitrogen. After cooling to room
temperature, ethyl acetate was added, and the organic phase
s, 3H), 2.47 (s, 3H), 2.96 (t, 2H), 3.60 (q, 2H), 5.52 (br s, 1H),
7
7
3
.02 (d, 1H, J ) 2.44 Hz), 7.05 (dd, 1H J ) 2.0 and 7.8 Hz),
-1
.29 (d, 1H), 7.39 (s, 1H), 7.99 (br s, 1H); IR (cm , neat): 3400,
292, 1652.
2 4
washed twice with water and dried (Na SO ). The solvent was
evaporated in vacuo, and the residue was purified by flash
column chromatography (silica gel, dichloromethane as eluent)
to give 0.42 g (60% yield) of the desired product as a yellow
N-Acetyl-5-ch lor otr yp ta m in e (AI-32). From 5-chlorot-
ryptamine:41 82% yield, mp 151 °C (EtOAc); lit mp 128-130;
30
+
1
1
EIMS: m/z 236 (M ), 177 (100); H NMR (CDCl ): δ 1.95 (s,
amorphous solid. H NMR (DMSO) δ 7.06-7.12 (ddd, 1H),
3
3
2
H), 2.93 (t, 2H), 3.57 (q, 2H), 5.59 (br s, 1H), 7.06 (d, 1H, J )
.44 Hz), 7.15 (dd, 1H, J ) 1.95 Hz and J ) 8.79 Hz), 7.30 (d,
7
.58-7.55 (q, 1H), 7.82-7.89 (dd, 1H), 8.06 (d, 1H, J ) 13.45
Hz), 8.29 (s, 1H), 8.39 (d, 1H, J ) 13.45 Hz), 12.31 (br s, 1H);
EIMS: m/z 206 (M ), 133 (100).
+
1H, J ) 8.79 Hz), 7.55 (d, 1H, J ) 1.95 Hz), 8.36 (br s, 1H); IR
-
1
(cm , Nujol): 3297, 3209, 1618.
Tr yp ta m in e Der iva tives: Gen er a l P r oced u r e. The
suitable 3-(2-nitroethenyl)-1H-indole (1 mmol) was added
portionwise to a stirred, ice-cooled suspension of LiAlH (0.23
4
g, 6 mmol) in dry THF (15 mL) under nitrogen, and the
mixture was stirred at room temperature for 5 h. After cooling
to 0 °C, water was added dropwise to destroy the excess
hydride, the mixture was filtered on Celite, and the filtrate
was concentrated in vacuo and partitioned between water and
ethyl acetate. The organic layer was washed with brine, dried
N-[2-[5-(2-Hyd r oxyeth yloxy)-1H-in d ol-3-yl]eth yl]a ceta -
4
2
m id e (AI-35). From 5-(hydroxyethyloxy)tryptamine: 70%
+
yield, mp 118 °C (dichloromethane); EIMS: m/z 262 (M ), 203
1
(
(
100); H NMR (acetone-d
6
): δ 1.86 (s, 3H), 2.87 (t, 2H), 3.46
q, 2H), 3.89 (m, 2H), 4.08 (t, 2H), 6.75 (dd, 1H, J ) 2.2 Hz
and J ) 8.79), 7.12 (d, 1H, J ) 2.2 Hz), 7.14 (d, 1H, J ) 2.2
-
1
Hz), 7.26 (d, 1H, J ) 8.79), 8.13 (br s, 1H); IR (cm , Nujol):
256, 3114, 1629. Anal. (C14 ) C, H, N.
N-Acetyl-6-br om o-5-m eth oxytr yp ta m in e (AI-38). From
6-bromo-5-methoxytryptamine:43 73% yield, mp 147-148 °C
3
18 2 3
H N O
2 4
(Na SO ), and concentrated under reduced pressure to give a
crude oily amine which was then used without further
purification. Chemical-physical data of these tryptamines
+
1
3
(CHCl /hexane); EIMS: m/z 310, 312 (M ), 251, 253 (100); H

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3841
7.09 (d, 1H, J ) 2.44 Hz), 7.19 (s, 1H), 7.33-7.49 (m, 6H); IR
NMR (acetone-d
6
): δ 1.85 (s, 3H), 2.87 (t, 2H), 3.46 (q, 2H),
-
1
3
.83 (s, 3H), 6.99 (d, 1H, J ) 2.44 Hz), 7.17 (d, 1H, J ) 1.95
20 2 2
(cm , Nujol): 3252, 1635. Anal. (C19H N O ) C, H, N.
-
1
P h a r m a cology. 2-[125I]-Iod om ela ton in Bin d in g Stu d -
ies a n d Liter a tu r e Affin ity Da ta . The affinity of the newly
synthesized aMT analogues for the melatonin receptor isolated
from quail optic tecta was determined in competition binding
Hz), 7.22 (d, 1H, J ) 2.44), 10.03 (br s, 1H); IR (cm , Nujol):
3
411, 3317, 1674. Anal. (C13
H
15BrN
2
O
2
‚0.02 CHCl
3
) C, H, N.
N-Acetyl-6-m eth oxytr yp ta m in e (AI-19). From 6-meth-
4
4
oxytryptamine: 86% yield, mp 137 °C (EtOAc/hexane), lit.
1
125
mp 136 °C; H NMR (CDCl
(
3
): δ 1.93 (s, 3H), 2.94 (t, 2H), 3.59
analyses using 2-[ I]-iodomelatonin as a labeled ligand (100
q, 2H), 3.86 (s, 3H), 5.52 (br s, 1H), 6.81 (dd, 1H, J ) 8.6 Hz
i
pM). The IC50 values were determined and K values calcu-
lated by nonlinear fitting (Table 2). The source of the animals,
the characterization of the melatonin receptor, and the isola-
and J ) 2.22 Hz), 6.88 (d, 1H, J ) 2.07 Hz), 6.94 (d, 1H, J )
1
2
.92 Hz), 7.47 (d, 1H, J ) 8.6 Hz), 7.95 (br s, 1H); EIMS: m/z
+
32 (M ), 173 (100).
tion of the crude membrane preparations have been described
in detail elsewhere.4
6,47
The affinity of compounds AI-3, -5,
2
,6-Dibr om om elaton in (AI-13). N-Bromosuccinimide (0.18
g, 1 mmol) was added portionwise to a solution of AI-38 (0.31
g, 1 mmol) in acetic acid (4 mL). The reaction mixture was
stirred under nitrogen at room temperature for 4 h and then
cooled at 0 °C, neutralized with a 50% solution of NaOH, and
extracted with ethyl acetate. The combined organic layers
were washed with brine, dried (Na SO ), and concentrated at
2 4
reduced pressure to give a crude residue which was purified
by flash chromatography (silica gel; ethyl acetate/cyclohexane,
-8, -9, -10, AII, BIII, CIII-1, DI, DII, and DIII, previously
1
25
synthesized by us, was also evaluated by displacing 2-[ I]-
iodomelatonin from quail brain membranes.
The affinity values of the other compounds examined in this
study were collected from literature (Table 1). To overcome
the problem of heterogeneity of binding data from different
laboratories, values were expressed as relative affinity (RA)
of the compounds tested compared to that of aMT in the same
experiment, and its negative logarithm (pRA) was used in the
PLS analyses as the dependent variable. When more binding
1
:1, as eluent) and crystallization from CHCl
g, 25%); mp 139-140 °C (CHCl ); EIMS: m/z 388, 390, 392
M ), 331 (100); H NMR (acetone-d ): δ 1.85 (s, 3H), 2.87 (t,
H), 3.39 (q, 2H), 3.9 (s, 3H), 7.18 (br s, 1H), 7.30 (s, 1H), 7.54
3
. Yield (0.098
3
+
1
(
2
6
i
constants were reported in the literature, we used the K values
obtained from chicken brain membranes, because they are the
most commonly used and because of the resemblance of chick
and quail brain melatonin receptors.46 The biological assays
on chicken brain membranes have been conducted with two
different experimental protocols, as reported by Sugden and
Chong48 or by Langlois et al.; we chose the data from the
former authors, since these are more numerous. The only
-
1
(
s, 1H), 8.02 (br s, 1H); IR (cm , Nujol): 3259, 3113, 1629.
Anal. (C13 ) C, H, N.
,4-Dibr om om ela ton in (AI-12) was obtained according to
H
2 2 2
14Br N O
2
the procedure described above starting from melatonin and
using 2 equiv of N-bromosuccinimide. 25% Yield; mp 177 °C
21a
+
(
acetone/hexane); EIMS: m/z 388, 390, 392 (M ), 318 (100);
1
exception was made for N-acyl-2-amino-8-methoxytetralin (BI-
H NMR (acetone-d
.86 (s, 3H), 6.99 (d, 1H, J ) 8.79 Hz), 7.20 (br s, 1H), 7.32 (d,
6
): δ 1.87 (s, 3H), 3.13 (t, 2H), 3.48 (q, 2H),
4c
3
1
3
1), whose K
i
value measured on chicken retina2 was used
-
1
instead of that on chicken brain membranes, to maintain the
same biological substrate as for the other tetralin derivatives.
For compounds BI-1, DIV-1, and DVI-1 the K values for
i
the two separated enantiomers have been reported, but we
decided to use the data referring to the racemic mixture, as
nothing is known about the absolute stereochemistry of DIV-1
and DVI-1. For BI-1 it is known that the most active
enantiomer is S(-),49 and for BI-3, BI-4, and DV-3 only the
affinity values for the separate enantiomers were available.
Since for other chiral compounds only the affinity of the
H, J ) 8.79 Hz), 10.82 (br s, 1H); IR (cm , Nujol): 3299,
177, 1610. Anal. (C13 ) C, H, N.
N -[(1-Be n zyl-5-m e t h oxy-1H -in d ol-3-yl)e t h yl]a ce t a -
m id e (AI-43). A solution of melatonin (1 mmol) in dry DMF
2 mL) was added dropwise to a stirred ice-cooled suspension
of sodium hydride (0.042 g of a 80% dispersion in mineral oil,
.4 mmol) in dry DMF (3 mL) under a N atmosphere. After
H
2 2 2
14Br N O
(
1
2
the addition, the mixture was stirred at 0 °C for 30 min, and
then benzyl chloride (0.15 mL, 1.3 mmol) was added dropwise
and the resulting mixture was stirred at room temperature
for 16 h and then poured into ice-water (25 g) and extracted
with ethyl acetate (3 × 10 mL). The organic phase was washed
with brine, dried over sodium sulfate, and concentrated under
reduced pressure to give a residue which was purified by
crystallization from EtOAc/hexane. Yield (0.25 g, 78%); mp
i
racemate was known, we derived the K values for the racemic
mixture as the harmonic mean of those of the enantiomers,
according to the method of Schaper.50
Not all the compounds included have been tested for
functional activity; among those tested, most were full agonists
at the melatonin receptor. Some of them have been reported
to show partial agonism on some pharmacological tests, but
the efficacy of these compounds strongly depends on tissue
preparations. Among the compounds reported in Table 1,
2-phenyl-aMT (AI-3) is a partial agonist on rabbit parietal
cortex model,26 but it is a full agonist on pigment aggregation
+
1
1
1
2
3
15 °C; EIMS: m/z 322 (M ), 91 (100); H NMR (CDCl ): δ
.92 (s, 3H), 2.95 (t, 2H), 3.58 (q, 2H), 3.86 (s, 3H), 5.25 (s,
H), 5.54 (br s, 1H), 6.86 (dd, 1H, J ) 2.55 Hz and J ) 8.9
-
1
Hz), 6.95 (s, 1H), 7.04-7.31 (m, 8H); IR (cm , Nujol): 3314,
1
641. Anal. (C20
22 2 2
H N O ) C, H, N.
N -[(5-Me t h oxy-1-m e t h yl-1H -in d ol-3-yl)e t h yl]a ce t a -
m id e (AI-42) was obtained according to the procedure de-
scribed above, starting from melatonin and using MeI as
5
1
in Xenopus laevis dermal melanophores. N-Acetyltryptamine
(AI-16) and its 2,6-dibromo derivative (AI-27) are partial
4
5
agonists on Xenopus melanophores.52 N-Acetyl-5-OH-tryp-
alkylating agent: 81% yield, mp 109 °C (EtOAc); lit. mp 100
+
1
tamine (AI-33) showed no agonist activity on Xenopus
melanophores,2 but its naphthyl analogue (AIII-7) was an
agonist on ovine pars tuberalis;2 similarly, N-propionyl-o-
methoxyphenylethylamine had no agonist activity on dopam-
ine release in rabbit retina, while its N-acetyl analogue is
reported to be an agonist.2 Other compounds endowed with
°
(
5
)
C; EIMS: m/z 246 (M ), 174 (100); H NMR (CDCl
s, 3H), 2.93 (t, 2H), 3.57 (q, 2H), 3.73 (s, 3H), 3.87 (s, 3H),
.63 (br s, 1H), 6.87 (s, 1H), 6.90 (dd, 1H, J ) 2.54 Hz and J
3
): δ 1.94
9d
1b
8.9 Hz), 7.03 (d, 1H, J ) 2.23 Hz), 7.21 (d, 1H, J ) 8.9 Hz);
IR (cm , Nujol): 3444, 1669.
-
1
4c
N -[(5-Me t h oxy-1-p h e n yl-1H -in d ol-3-yl)e t h yl]a ce t a -
2
2
22
24d
24b
partial agonism were AII-7, AII-8, BI-10, BIII-1, CI-
m id e (AI-44). A mixture of melatonin (0.23 g, 1 mmol), K
CO (0.175 g, 1.27 mmol), iodobenzene (0.35 g, 1.73 mmol),
CuI (0.05 g), and ZnO (0.012 g) in 1-methyl-2-pyrrolidinone
NMP) (2 mL) was heated at 155 °C for 6 h. After cooling to
°C, the salts were filtered, and the filtrate was partitioned
between Et O and 2 N NH OH. The organic phase was
2
-
2
9d
24b
24a
1
5, DI-2, DIII-6.
3
Da ta Set a n d Cla ssifica tion . One hundred thirty-three
(
0
compounds were taken into account: they share a common
N-acetylamino or N-propionylamino group in the side chain,
but differ as to both the nature of the aryl moiety and the
length of the side chain. Despite this structural heterogeneity,
it was possible to identify three different groups which share
common characteristics.
Subset A: melatonin-like compounds (AI-1-AI-17, AI-21-
AI-46, AII-1-AII-6, AIII-1, -2, -6, -7, AIV). These compounds,
identical from a topological point of view to aMT, have an
2
4
washed with brine, dried over sodium sulfate, and concen-
trated under reduced pressure to give a residue which was
purified by crystallization from ethyl acetate: yield (0.23 g,
+
1
7
4.6%); mp 119 °C; EIMS: m/z 308 (M ), 236 (100); H NMR
CDCl ): δ 1.97 (s, 3H), 3.0 (t, 2H), 3.64 (q, 2H), 3.90 (s, 3H),
.65 (br s, 1H), 6.91 (dd, 1H, J ) 2.44 Hz and J ) 8.78 Hz),
(
3
5

3
842 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
N-acylaminoethyl side chain bound to an aromatic nucleus,
i.e., an indole or a naphthalene. The indole derivatives can
have the amidoethyl side chain either in position 3, as in the
natural ligand, or in position 1, as in the recently reported
derivatives (AII).2 Compounds lacking the topological equiva-
lence (relative position of methoxy group and side chain) with
aMT for the methoxy position (AI-18-AI-20, AII-7, AII-8,
AIII-5) were excluded from subset A and included in the global
set. Compounds AIII-3 and AIII-4 had two possible align-
ments which are discussed in the text.
Subset B: N-acyl-2-amino-8-methoxytetralins and related
compounds (BI-1-BI-6, BI-9, -10, BII, BIII, BIV). In this
conformationally constrained set of compounds the melatonin-
like ethyl side chain is part of a six atom ring, with the
N-acylamino group directly bound to this cycle. Again,
compounds lacking the topological equivalence regarding the
methoxy group position (BI-7 and BI-8) were excluded.
their values of potency are closer to those of aMT derivatives
than of N-acylphenylalkylamines. This assignment was con-
firmed by the prediction of the affinity in the two orientations
by the global 3D-QSAR model (see Results and Discussion).
2
Among the N-acylphenylalkylamines in subset C the es-
sential methoxy group can be in ortho or meta to the alkyl
chain. The two compounds having both an o- and m-methoxy
group (CI-3, -4) were aligned, superposing the ortho group onto
that of aMT because, from qualitative SAR, it appears that
with the acylaminoethyl chain, o-methoxy derivatives are
generally more potent than m-methoxy ones.
En er gy Min im iza tion a n d Con for m er Selection . Mo-
lecular modeling studies were performed with the SYBYL 6.3
33
Software running on a Silicon Graphics R4400 200 MHz 64
Mb RAM Indigo2 workstation. Three-dimensional models of
all molecules were built and energy minimized using the
standard Tripos force field,53 excluding the electrostatic con-
Subset C: N-acylphenylalkylamines and related compounds
54
tribution, with the Powell method and a convergence gradient
(
CI, CII, CIII-1), in which the methoxy group and the
of 0.02 kcal/mol‚Å. For the seven constrained molecules (BI-
N-acylaminoalkyl side chain originate from the same ring. This
set includes not only phenyl derivatives, but also naphthyl or
indole ones, characterized by the presence of the alkoxy group
in the same benzene ring as the acylaminoalkyl chain (CII,
CIII-1). In some phenyl derivatives (CI-19-CI-23) the meth-
oxy group was replaced by a halogen atom.
1
, BIII-1, DI-1, DII-1, DIII-1, DIV-1, DV-1) used to derive
the pharmacophore models (see above), the minimum energy
conformations proposed by the best overlap option of DISCO
24b
were used; the molecules structurally related to them were
energy minimized in the corresponding local minima. For the
N-acyl-2-amino-8-methoxytetralins and related compounds
Each subset was considered separately and a 3D-QSAR
model was devised for each one in order to clarify the
structure-affinity relationships for each homogeneous class
of derivatives. The remaining compounds, referred to as
subset D for the sake of clarity, could not be included in the
previously cited subsets A, B, and C because they are different
from a topological or structural point of view; they were
included in the global model obtained for the melatonin
receptor ligands.
(
subset B), the spatial configurations corresponding to the
S-enantiomer of compound BI-1 (see above) were considered.
For the remaining compounds, the conformers which best
fitted the aMT spatial geometry were chosen. The molecules
were aligned by means of a rigid body fitting procedure,
superposing the methoxy oxygen (when present), the four
atoms in the amido group, and the six atoms of the aryl moiety
to those of aMT in the putative active conformations as
suggested by the pharmacophore models A and B.
Molecu la r Mod elin g. Meth oxy Gr ou p Or ien ta tion .
The aMT methoxy group appears free to rotate, with a low
energy barrier, as confirmed by AM1 calculations (MOPAC 6.0,
CoMF A. For structure-affinity studies the QSAR CoMFA
56
module of SYBYL was used, calculating the steric and elec-
trostatic fields within a lattice with a grid resolution of 1 Å,
whose extension was at least 4 Å beyond every molecule
boundary in all directions: an sp carbon with a point charge
of 1.0 was taken as the probe atom. The electrostatic field
was calculated from Gasteiger-H u¨ ckel charges, with the
dielectric function depending on 1/r. For those points where
the steric cutoff (30 kcal/mol) was reached, the electrostatic
potential was in practice excluded from the analyses by fixing
3
3
implemented in SYBYL ) on 36 rotamers of 5-methoxyindole
with torsion angles around the C5-O bond differing by 10°:
two minimum energy conformations were found at 0° and 180°,
with a rotational barrier of 0.6 kcal/mol. We decided to keep
the methoxy group in the plane of the indole ring, with the
methyl group pointing toward the side chain, to reproduce the
orientation of the rather potent constrained naphthopyran and
naphthofuran derivatives (AIV). The reliability of this orien-
tation is confirmed by those compounds which cannot assume
it (AI-12, -15, BI, BII, BIII, BIV, DII, DV, DVI-1, for
example), which all exhibit less affinity than aMT. This lack
of potency could be due to a conformational effect of the
substituents on the orientation of the side chain, when free to
rotate, and/or to an unexpected disposition of the methoxy
group. For those compounds which, unlike aMT, cannot orient
the methoxy group, we decided to fix it in the nearest energy
minimum conformation, that is, with the methyl group per-
pendicular to the indole nucleus.
3
5
5
it to the mean of all the nonexcluded electrostatic values
calculated in the same grid point.3
1,57
In addition, a lipophi-
3
2
licity field (MLP) was calculated using the CLIP program.
All regression analyses were performed using the Partial Least
Squares (PLS)58 algorithms in SYBYL; different combinations
of the three field descriptors were taken into account (steric
only, lipophilic only, steric-electrostatic, steric-lipophilic,
steric-electrostatic-lipophilic) in order to verify their correla-
tion with the dependent variable and the possible presence of
intercorrelation.
The optimal number of latent variables was chosen by
means of the cross-validation technique,59 using the leave-one-
out procedure; only variables with an energy standard devia-
tion higher than 2 kcal/mol were included in the cross-
validated runs, to reduce the computation time and to minimize
the influence of noisy columns. The predictive power of the
model was also tested excluding one-fourth of the set com-
pounds, randomly chosen, in each cross-validation run (four
cross-validation groups). We observed that both the optimal
number of latent variables and Q2 values were stable with
respect to the cross-validation method employed (leave one out
or four groups). Moreover, we observed that the exclusion of
the variables with standard deviation lower than 2 kcal/mol
Align m en t Ru les. Compounds were initially aligned on
the putative active conformations of aMT, obtained from our
previous pharmacophore search (model A: τ
CR) ) 77.6°; τ (C3-Câ-CR-N) ) -179.8°; τ
C) ) 79.4°; model B: τ ) 72.1°, τ ) -179.8°, τ
which led us to propose two different pharmacophore models.
1
(C3a-C3-Câ-
(Câ-CR-N-
) -179.2°),
2
3
1
2
3
2
4b
The atoms used for the alignment were those presumed to
interact with the receptor: referring to aMT, the methoxy
oxygen, the phenyl ring of the indole nucleus, and the amido
moiety. Some compounds showed different possibilities of
alignment. Naphthalene compounds CII lack the methoxy
group in position 7, topologically equivalent to aMT position
5
, but have an alkoxy group ortho to the side chain; they were
2
included in subset C, as we considered the methoxy group
interaction to be more important than that occurring at
position 2 of aMT derivatives. This hypothesis of alignment
brought no change in the Q value on subset A with four cross-
validation groups (differences < 0.02 kcal/mol in models with
1-6 latent variables, data not shown). The final non-cross-
validated PLS analyses were derived, with no energy filtering
applied, for the combination of fields and the number of latent
2
1a
has already been proposed by Langlois et al.
Compounds
AIII-3, -4 have two possible alignments, that of subset A or
subset C; they were considered as members of subset A, as
2
variables giving the highest Q value.

Melatonin Receptor Ligands
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20 3843
Ack n ow led gm en t. Financial support from the Ital-
ian MURST (40% and 60%) and the CNR is gratefully
acknowledged.
(17) Giusti, P.; Lipartiti, M.; Franceschini, D.; Schiavo, N.; Floreani,
M.; Manev, H. Neuroprotection by Melatonin from Kainate-
Induced Exitotoxicity in Rats. FASEB J . 1996, 10, 891-896.
(18) Pozo, D.; Reiter, R. J .; Calvo, J . R.; Guerrero, J . M. Physiological
Concentrations of Melatonin Inhibit Nitric Oxide Synthase in
Rat Cerebellum. Life Sci. 1994, 55, 455-460.
Refer en ces
(
19) (a) Lipartiti, M.; Franceschini, D.; Zanoni, R.; Gusella, M.; Giusti,
P.; Cagnoli, C. M.; Kharlamov, A.; Manev, H. Neuroprotective
Effects of Melatonin. Adv. Exp. Med. Biol. 1996, 398, 315-321.
(
1) Reiter, R. J . Pineal Melatonin: Cell Biology of its Synthesis and
of its Physiological Interactions. Endocr. Rev. 1991, 12, 151-
(b) Manev, H.; Uz, T.; Kharlamov, A.; J oo, J . Y. Increased Brain
1
80.
Damage after Stroke or Excitotoxic Seizures in Melatonin-
Deficient Rats. FASEB J . 1996, 10, 1546-1551. (c) Bertuglia,
S.; Marchiafava, P. L.; Colantuoni, A. Melatonin Prevents
Ischemia Reperfusion Injury in Hamster Cheek Pouch Micro-
circulation. Cardiovasc. Res. 1996, 31, 947-952.
(
2) (a) Reiter, R. J . The Melatonin Rhythm: both a Clock and a
Calendar. Experientia 1993, 49, 654-664. (b) Underwood, H. The
Pineal and Melatonin: Regulators of Circadian Function in
Lower Vertebrates. Experientia 1990, 46, 120-128.
3) Cahill, G. M.; Besharse, J . C. Circadian Rhythmicity in Verte-
brate Retinas: Regulation by a Photoreceptor Oscillator. Progress
in Retinal Eye Research; Elsevier Science Ltd.: Great Britain,
(
(
(
20) Takaki, K. S.; Mahle, C. D.; Watson A. J . Melatonergic
Ligands: Pharmaceutical Development and Clinical Application.
Curr. Pharm. Des. 1997, 3, 429-438.
1
995; Vol. 14, pp 267-291.
4) Tamarkin, L.; Baird, C. J .; Almeida, O. F. X. Melatonin:
Coordinating Signal for Mammalian Reproduction? Science
985, 227, 714-720.
21) (a) Langlois, M.; Br e´ mont, B.; Shen, S.; Poncet, A.; Andrieux,
J .; Sicsic, S.; Serraz, I.; Math e´ -Allainmat, M.; Renard, P.;
Delagrange, P. Design and Synthesis of New Naphthalenic
(
(
a
1
1
25
Derivatives as Ligands for 2-( I)-Iodomelatonin Binding Sites.
J . Med. Chem. 1995, 38, 2050-2060. (b) Depreux, P.; Lesieur,
D.; Mansour, H. A.; Morgan, P.; Howell, H. E.; Renard, P.;
Caignard, D. H.; Pfeiffer, B.; Delagrange, P.; Guardiola, B.; Yous,
S.; Demarque, A.; Adam, G.; Andrieux, J . Synthesis and Struc-
ture-Activity Relationships of Novel Naphthalenic and Bioi-
sosteric Related Amidic Derivatives as Melatonin Receptor
Ligands. J . Med. Chem. 1994, 37, 3231-3239. (c) Yous, S.;
Andrieux, J .; Howell, H. E.; Morgan, P. J .; Renard, P.; Pfeiffer,
B.; Lesieur, D.; Guardiola-Lemaitre, B. Novel Naphthalenic
Ligands with High Affinity for the Melatonin Receptor. J . Med.
Chem. 1992, 35, 1484-1486.
5) Oldani, A.; Ferini-Strambi, L.; Zucconi, M.; Stankov, B.; Fras-
chini, F.; Smirne, S. Melatonin and Delayed Sleep Phase
Syndrome: Ambulatory Polygraphic Evaluation. NeuroReport
1
994, 6, 132-134.
(
6) Rosenthal, N. E.; Sack, D. A.; J acobsen, F. M.; J ames, S. P.;
Parry, B. L.; Arendt, J .; Tamarkin, L.; Wehr, T. A. Melatonin in
Seasonal Affective Disorder and Phototherapy. Neural Transm.
Suppl. 1986, 21, 257-267.
7) (a) Petrie, K.; Conaglen, J . V.; Thompson, L.; Chamberlain, K.
Effect of Melatonin on J et Lag after Long Haul Flights. Br. Med.
J . 1989, 298, 705-707. (b) Arendt, J .; Aldhous, M.; Marks, V.
Alleviation of J et Lag by Melatonin: Preliminary Results of a
Controlled Double-Blind Trial. Annu. Rev. Chronopharmacol.
(
(22) Tarzia, G.; Diamantini, G.; Di Giacomo, B.; Spadoni, G.; Esposti,
D.; Nonno, R.; Lucini, V.; Pannacci, M.; Fraschini, F.; Stankov,
1
986, 3, 49-52.
B. M. 1-(2-Alkanamidoethyl)-6-methoxyindole Derivatives:
New Class of Potent Indole Melatonin Analogues. J . Med. Chem.
997, 40, 2003-2010.
a
(
8) Sack, R. L.; Blood, M. L.; Lewy, A. J . Melatonin Rhythms in
Night Shift Workers. Sleep 1992, 15, 434-441.
9) (a) Garfinkel, D.; Laudon, M.; Zisapel, N. Improvement of Sleep
Quality by Controlled-Release Melatonin in Benzodiazepine-
Treated Elderly Insomniacs. Arch. Gerontol. Geriatr. 1997, 24,
1
(
(
23) Garratt, P. J .; Travard, S.; Vonhoff, S.; Tsotinis, A.; Sugden, D.
Mapping the Melatonin Receptor. 4. Comparison of the Binding
Affinities of a Series of Substituted Phenylalkyl Amides. J . Med.
Chem. 1996, 39, 1797-1805.
2
23-231. (b) Sack, R. L.; Lewy, A. J .; Parrott, K.; Singer, C. M.;
McArthur, A. J .; Blood, M. L.; Bauer, V. K. Melatonin Analogues
(
24) (a) Garratt, P. J .; Vonhoff, S.; Rowe, S. J .; Sugden, D. Mapping
the Melatonin Receptor. 2. Synthesis and Biological Activity of
Indole Derived Melatonin Analogues with Restricted Conforma-
tions of the C-3 Amidoethane Side Chain. Bioorg. Med. Chem.
Lett. 1994, 4, 1559-1564. (b) Spadoni, G.; Balsamini, C.;
Diamantini, G.; Di Giacomo, B.; Tarzia, G.; Mor, M.; Plazzi, P.
V.; Rivara, S.; Lucini, V.; Nonno, R.; Pannacci, M.; Fraschini,
F.; Stankov, B. M. Conformationally Restrained Melatonin
Analogues: Synthesis, Binding Affinity for the Melatonin Recep-
tor, Evaluation of the Biological Activity, and Molecular Model-
ing Study. J . Med. Chem. 1997, 40, 1990-2002. (c) Copinga, S.;
Tepper, P. G.; Grol, C. J .; Horn, A. S.; Dubocovich, M. L.
and Circadian Sleep Disorders. Eur. J . Med. Chem. 1995, 30,
6
61-669s. (c) Zhdanova, I. V.; Wurtman, R. J .; Lynch, H. J .;
Ives, J . R.; Dollins, A. B.; Morabito, C.; Matheson, J . K.; Schomer,
D. L. Sleep-Inducing Effects of Low Doses of Melatonin Ingested
in the Evening. Clin. Pharmacol. Ther. (St. Louis) 1995, 57,
5
52-558.
(
(
(
10) Morgan, P. J .; Lawson, W.; Davidson, G.; Howell, H. E. Guanine
Nucleotides Regulate the Affinity of Melatonin Receptors on the
Ovine Pars tuberalis. Neuroendocrinology 1989, 50, 359-362.
11) Reppert, S. M.; Weaver, D. R.; Godson, C. Melatonin Receptors
Step into the Light: Cloning and Classification of Subtypes.
Trends Pharmacol. Sci. 1996, 17, 100-102.
12) Ebisawa, T.; Karne, S.; Lerner, M. R.; Reppert, S. M. Expression
Cloning of a High-Affinity Melatonin Receptor from Xenopus
Dermal Melanophores. Proc. Natl. Acad. Sci. U.S.A. 1994, 91,
2
-Amido-8-Methoxytetralins: A Series of Nonindolic Melatonin-
Like Agents. J . Med. Chem. 1993, 36, 2891-2898. (d) Sugden,
D. N-Acyl-3-amino-5-methoxychromans: a New Series of Non-
Indolic Melatonin Analogues. Eur. J . Pharmacol. 1994, 254,
6
133-6137.
2
71-275. (e) Leclerc, V.; Depreux, P.; Lesieur, D.; Caignard, D.
(
13) (a) Reppert, S. M.; Weaver, D. R.; Ebisawa, T. Cloning and
Characterization of a Mammalian Melatonin Receptor That
Mediates Reproductive and Circadian Responses. Neuron 1994,
H.; Renard, P.; Delagrange, P.; Guardiola-Lemaitre, B.; Morgan,
P. Synthesis and Biological Activity of Conformationally Re-
stricted Tricyclic Analogues of the Hormone Melatonin. Bioorg.
Med. Chem. Lett. 1996, 6, 1071-1076. (f) Math e´ -Allainmat, M.;
Gaudy, F.; Sicsic, S.; Dangy-Caye, A. L.; Shuren, S.; Br e´ mont,
B.; Benatalah, Z.; Langlois, M.; Renard, P.; Delagrange, P.
Synthesis of 2-Amido-2,3-Dihydro-1H-Phenalene Derivatives as
New Conformationally Restricted Ligands for Melatonin Recep-
tor. J . Med. Chem. 1996, 39, 3089-3095. (g) Gruppen, G.; Grol,
C. J . Synthesis of New Melatonin Agonists. 10th Camerino-
Noordwijkerhout symposium: perspectives in receptor research
(Camerino, Italy, September 10-14, 1995).
1
3, 1177-1185. (b) Reppert, S. M.; Godson, C.; Mahle C. D.;
Weaver, D. R.; Slaugenhaupt, S. A.; Gusella, J . F. Molecular
Characterization of a Second Melatonin Receptor Expressed in
Human Retina and Brain: The Mel1b Melatonin Receptor. Proc.
Natl. Acad. Sci. U.S.A. 1995, 92, 8734-8738. (c) Mazzucchelli,
C.; Pannacci, M.; Nonno, R.; Lucini, V.; Fraschini, F.; Stankov,
B. M. The Melatonin Receptor in the Human Brain: Cloning
Experiments and Distribution Studies. Mol. Brain Res. 1996,
3
9, 117-126.
(
14) (a) Maestroni, G. J . M. The Immunoneuroendocrine Role of
Melatonin. J . Pineal Res. 1993, 14, 1-10. (b) Lissoni, P.; Barni,
S.; Tancini, G.; Ardizzoia, A.; Cazzaniga, M.; Frigerio, F.; Brivio,
F.; Conti, A.; Maestroni, G. J . M. Neuroimmunomodulation of
Interleukin-2 Cancer Immunotherapy by Melatonin: Biological
and Therapeutic Results. Adv. Pineal Res. 1994, 7, 183-189.
(25) Sugden, D.; Chong, N. W. S.; Lewis, D. F. V. Structural
Requirements at the Melatonin Receptor. Br. J . Pharmacol.
1995, 114, 618-623.
(26) Spadoni, G.; Stankov, B.; Duranti, A.; Biella, G.; Lucini, V.;
Salvatori, A.; Fraschini, F. 2-Substituted 5-Methoxy-N-Acyl-
tryptamines: Synthesis, Binding Affinity for the Melatonin
Receptor, and Evaluation of the Biological Activity. J . Med.
Chem. 1993, 36, 4069-4074.
(27) Martin, Y. C.; Bures, M. G.; Danaher, E. A.; De Lazzer, J .; Lico,
I.; Pavlik, P. A. A Fast New Approach to Pharmacophore
Mapping and its Application to Dopaminergic and Benzodiaz-
epine Agonists. J . Comput.-Aided Mol. Des. 1993, 7, 83-102.
(28) Sicsic, S.; Serraz, I.; Andrieux, J .; Br e´ mont, B.; Math e´ -Allainmat,
M.; Poncet, A.; Shen, S.; Langlois, M. Three-Dimensional
Quantitative Structure-Activity Relationship of Melatonin Re-
ceptor Ligands: A Comparative Molecular Field Analysis Study.
J . Med. Chem. 1997, 40, 739-748.
(
c) Maestroni, G. J . M.; Georges, J . M.; Hertens, E.; Galli, P.;
Conti, A.; Pedrinis, E. Melatonin-Induced T-Helper Cell He-
matopoietic Cytokines Resembling both Interleukin-4 and Dynor-
phin. J . Pineal Res. 1996, 21, 131-139.
(
15) Lissoni, P.; Paolorossi, F.; Tancini, G.; Ardizzoia, A.; Barni, S.;
Brivio, F.; Maestroni, G. J . M.; Chilelli, M. A Phase II Study of
Tamoxifen Plus Melatonin in Metastatic Solid Tumor Patients.
Br. J . Cancer 1996, 74, 1466-1468.
16) Reiter, R. J . The Indoleamine Melatonin as a Free Radical
Scavenger, Electron Donor, and Antioxidant: in Vitro and in
Vivo Studies. Adv. Exp. Med. Biol. 1996, 398, 307-313.
(

3
844 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 20
Mor et al.
(
29) (a) Grol, C. J .; J ansen, J . M. The High Affinity Melatonin Binding
Site Probed with Conformationally Restricted Ligands-II. Ho-
mology Modeling of the Receptor. Bioorg. Med. Chem. 1996, 4,
(45) Frohn, M. A.; Seaborn, C. J .; J ohnson, D. W.; Phillipou, G.;
Seamark, R. F.; Matthews, C. D. Structure-Activity Relation-
ship of Melatonin Analogues. Life Sci. 1980, 27, 2043-2046.
46) Cozzi, B.; Stankov, B.; Viglietti-Panzica, C.; Capsoni, S.; Aste,
N.; Lucini, V.; Fraschini, F.; Panzica, G. C. Distribution and
Characterization of Melatonin Receptor in the Brain of the
J apanese Quail, Coturnix J aponica. Neurosci. Lett. 1993, 150,
1
333-1339. (b) J ansen, J . M.; Copinga, S.; Gruppen, G.; Moli-
(
nari, E. J .; Dubocovich, M. L.; Grol, C. J . The High Affinity
Melatonin Binding Site Probed with Conformationally Restricted
Ligands-I. Pharmacophore and Minireceptor Models. Bioorg.
Med. Chem. 1996, 4, 1321-1332. (c) Reppert, S. M.; Weaver D.
R.; Cassone V. M.; Godson, C.; Kolakowski, L. F., J r. Melatonin
Receptor are for the birds - Molecular Analysis of Two Receptor
subtypes Differentially expressed in Chick brain. Neuron 1995,
1
49-152.
(
47) Stankov, B.; Cozzi, B.; Lucini, V.; Fumagalli, P.; Scaglione, F.;
Fraschini, F. Characterization and Mapping of Melatonin Recep-
tors in the Brain of Three Mammalian Species: Rabbit, Horse
and Sheep. A Comparative in Vitro Binding Study. Neuroendo-
crinology 1991, 53, 214-221.
1
5, 1003-1015. (d) Pickering, H.; Sword, S.; Vonhoff, S.; J ones,
R.; Sugden D. Analogues of Diverse Structure are Unable to
Differentiate Native Melatonin Receptors in the Chicken Retina,
Sheep Pars tuberalis and Xenopus Melanophores. Br. J . Phar-
macol. 1996, 119, 379-387.
(48) Sugden, D.; Chong, N. W. S. Pharmacological Identity of 2-[125I]-
Iodomelatonin Binding Sites in Chicken Brain and Sheep Pars
tuberalis. Brain Res. 1991, 539, 151-154.
49) Copinga, S. The Semirigid 2-Aminotetralin System: A Structural
Base for Dopamine- and Melatonin-Receptor Agents. Thesis
defended at the State University of Groningen, The Netherlands,
Faculty of Science, October 1994.
(
(
(
30) Navajas, C.; Kokkola, T.; Poso, A.; Honka, N.; Gynther, J .;
Laitinen, J . T. A Rhodopsin-Based Model for Melatonin Recogni-
tion at its G Protein-Coupled Receptor. Eur. J . Pharmacol. 1996,
(
3
04, 173-183.
31) Cramer, R. D., III.; Patterson, D. E.; Bunce, J . D. Comparative
Molecular Field Analysis (CoMFA). 1. Effect of Shape on Binding
of Steroids to Carrier Proteins. J . Am. Chem. Soc. 1988, 110,
(
50) Schaper, K.-J . QSAR Analysis of Chiral Compounds Including
Racemates. In Progress in clinical and biological research.
QSAR: Quantitative Structure-Activity Relationships in Drug
Design, vol. 291; Fauch e` re, J . L., Ed.; Alan R. Liss, Inc.: New
York, 1989; pp 41-44.
5
959-5967.
32) Gaillard, P.; Carrupt, P.-A.; Testa, B.; Boudon, A. Molecular
Lipophilicity Potential, a Tool in 3D QSAR: Method and
Applications. J . Comput.-Aided Mol. Des. 1994, 8, 83-96.
33) SYBYL Molecular Modeling Software ver. 6.3, Tripos Inc., St.
Louis (MO), USA.
(
(
(
51) Garratt, P. J .; J ones, R.; Tocher, D. A.; Sugden, D. Mapping the
Melatonin Receptor. 3. Design and Synthesis of Melatonin
Agonists and Antagonists Derived from 2-Phenyltryptamines.
J . Med. Chem. 1995, 38, 1132-1139.
34) Duranti, E.; Stankov, B.; Spadoni, G.; Duranti, A.; Lucini, V.;
Capsoni, S.; Biella, G.; Fraschini, F. 2-Bromomelatonin: Syn-
thesis and Characterization of a Potent Melatonin Agonist. Life
Sci. 1992, 51, 479-485.
35) Still, I. W. J .; Strautmanis, J . R. Approaches to the Tetracyclic
Eudistomins: the Synthesis of N(10)-Acetyleudistomin L. Can.
J . Chem. 1990, 68, 1408-1419.
36) Yang, S. W.; Cordell, G. A. Metabolism Studies of Indole
Derivatives Using a Staurosporine Producer, Streptomyces
Staurosporeus. J . Nat. Prod. 1997, 60, 44-48.
(
52) Garratt, P. J .; J ones, R.; Rowe, S. J .; Sugden, D. Mapping the
Melatonin Receptor. 1. The 5-Methoxyl Group of Melatonin Is
Not an Essential Requirement for Biological Activity. Bioorg.
Med. Chem. Lett. 1994, 4, 1555-1558.
(
(
(
(53) (a) SYBYL 6.3 Force Field Manual, p 234, Tripos Inc., St. Louis,
MO. (b) Clark, M.; Cramer, R. D., III.; Van Opdenbosch, N.
Validation of the General Purpose Tripos 5.2 Force Field. J .
Comput. Chem. 1989, 10, 982-1012.
37) Kennaway, D. J .; Hugel, H. M.; Clarke, S.; Tjandra, A.; J ohnson,
D. W.; Royles, P.; Webb, H. A.; Carbone, F. Structure-Activity
Studies of Melatonin Analogues in Prepubertal Male Rats. Aust.
J . Biol. Sci. 1988, 41, 393-400.
(
54) Powell, M. J . D. Restart Procedures for the Conjugate Gradient
Methodol. Mathematical Programming 1977, 12, 241-254.
55) SYBYL 6.3 Force Field Manual, p 290, Tripos Inc., St Louis, MO.
56) SYBYL 6.3 Ligand-Based Design Manual; Tripos Inc., St. Louis,
MO; pp 215-236.
(
(
(
38) Velluz, L.; Muller, G.; J oly, R.; Nomine, G.; Mathieu, J .; Allais,
A.; Warnant, J .; Valls, J .; Bucourt, R.; J olly, J . Synthesis of
Reserpine and New Derivatives of Yohimban. Bull. Soc. Chim.
Fr. 1958, 673-677.
39) Yamada, F.; Saida, Y.; Somei, M. Structural Determination of a
Natural Alkaloid, 5-Methoxy-1-oxo-1,2,3,4-tetrahydro-â-carbo-
line and the Synthesis of the Corresponding 8-Methoxy Com-
pound. Heterocycles 1986, 24, 2619-2627.
(
57) Cramer, R. D., III.; DePriest, S. A.; Patterson, D. E.; Hecht, P.
The Developing Practice of Comparative Molecular Field Analy-
sis. In 3D QSAR in Drug Design. Theory Methods and Applica-
tions; Kubinyi, H., Ed.; ESCOM: Leiden, 1993; p 456.
(
(58) Wold, S.; Ruhe, A.; Wold, H.; Dunn, W. J . The Covariance
Problem in Linear Regression. The Partial Least Squares (PLS)
Approach to Generalized Inverses. SIAM J . Sci. Stat. Comput.
(
(
(
40) Crohare, R.; Merkuza, V. M.; Gonzalez, H. A.; Ruveda, E. A. 5,7-
Dimethoxyindole and Related Compounds. J . Heterocycl. Chem.
1
970, 7, 729-732.
1
984, 5, 735-743.
41) Buzas, A.; Herisson, C.; Lavielle, G. Application of the Wittig-
Horner Reaction to Indolinones. A Convenient Synthesis of
Tryptamines. Synthesis 1977, 129-130.
(
59) Cramer, R. D., III.; Bunce, J . D.; Patterson, D. E. Crossvalida-
tion, Bootstrapping, and Partial Least Squares Compared with
Multiple Regression in Conventional QSAR Studies. Quantum
Struct.-Act. Relat. 1988, 7, 18-25.
60) Dubocovich, M. L. Melatonin Receptors: Are There Multiple
Subtypes? Trends Pharmacol. Sci. 1995, 16, 50-56.
42) Suvorov, N. N.; Gordeev, E. N.; Vasin, M. V. Indole derivatives.
CI. Synthesis and Biological Activity of Some Tryptamines.
Khim. Geterotsikl. Soedin. 1974, 11, 1496-1501. [Chem. Abstr.
(
1
975, 82, 139888g].
(
43) Hino, T.; Lai, Z.; Seki, H.; Hara, R.; Kuramochi, T.; Nakagawa,
M. 1-(1-Pyrrolin-2-yl)-â-carbolines. Synthesis of Eudistomins H,
I, and P. Chem. Pharm. Bull. 1989, 37, 2596-2600.
(61) Cruciani, G.; Clementi, S., Baroni, M. Variable Selection in PLS
Analysis. In 3D QSAR in Drug Design. Theory Methods and
Applications; Kubinyi, H., Ed.; ESCOM: Leiden, 1993; p 552.
(
44) Spath, E.; Lederer, E. Synthese der Harmala-Alkaloide: Har-
malin, Harmin und Harman. Ber. 1930, 63B, 120-125.
J M9810093